Presented at 5G World 2016
In this video we share the complete recording of the SysMech and Vodafone Ireland Case study, presented at 5G World 2016. SysMech's Head of Product, Mark Slinger is joined by Michael Mullen, Manager of Network Optimisation at Vodafone to share the Vodafone Ireland story. The pair discuss the importance of end-to-end operational intelligence as the telecoms industry rapidly evolves.
Michael then shares the Vodafone Ireland story, explaining their operational intelligence goals, and the steps they have taken with Zen to achieve these over the past 12 months. He also shares a set of real data analytics dashboards, giving viewers a true insight into operational intelligence at Vodafone Ireland.
Mark Slinger: Good afternoon everyone. I’m Mark from SysMech and I’ll be giving you an introduction to what we have been doing with Vodafone Ireland around operational intelligence, and then I will hand you over to Michael who will talk a little bit about their journey.
First off, just to put this in context, what am I actually talking about, operational intelligence? Put simply, we are talking about the ability to make intelligent decisions based on near real-time data so that you can actually get insight and visibility into the network, make decisions, and do something intelligent with that data. In our experience, typically across the service provider community this has been done in a very silo’d approach. You’ve got different tools for different parts of the network, whether that’s from the radio access, the transmission, the core, the IMS. You’ve got different tools for different domains, such as performance management, the fault management space, customer experience management and so on. And you’ve also got different tools for different teams; typically we see lots of different tools across optimisation, operations, care and so on. But really in today’s competitive market, operators need to be able to act very quickly - you need to be able to troubleshoot problems quickly, you need to be able to understand the impact of issues, and you need to be able to identify new opportunities as well. So this kind of silo’d approach is no longer viable, in today’s market.
The only way to do this is to use a correlated view across the network to understand what is going on. But again that comes with its own complexities; with the sheer complexity, and sheer amounts of data across the different parts of the network, whether that’s from the mobile, or the fixed or the Wi-Fi part of the network, there is just so much data to be able to deal with. So how does somebody actually do that with today’s processes of looking at lots of different tools across the business?
The only way to do this is for operational intelligence to evolve towards an automatically correlated view across the whole network, the IT and the customer information, so you can get a real holistic understanding of what’s going on across the whole picture.
A good example here is some of the projects we have been involved in, where you look at network virtualisation; being able to ensure service continuity between moving applications from the physical world to the virtual world.
One of the other things that has become clear as well is that actually, being able to deal with so much data, visualisation is great but actually, automation and future prediction are no longer now ‘nice to haves,’ instead they are mandatory to be able to actually run your network as well.
I don’t want to talk too much about our product, if you are interested, come to our stand. In short we are working with communications providers to help them achieve those goals, by actually putting real-time, digestible information into the hands of hundreds of users, being able to automate that activity, and being able to ensure that the platform itself is easy to manage, so you don’t need teams of people to be able to manage that too.
So what I want to do now is to hand over to Michael from Vodafone who will give you a little bit of an insight into the customer journey.
Michael Mullen: Thanks Mark. Hi, I’m Michael Mullen, I’m the optimisation manage at Vodafone. I’d prefer if you call me Mick if you are going to come to me afterwards, in Ireland we always do thing a little bit shorter. I’m going to talk to you about operational intelligence in Ireland and the challenges that we met and faced, and how Zen and SysMech have helped us overcome those.
So what is operational intelligence, and what are the challenges that we have faced? So essentially what you have, is all these different facets of information and data sources coming from different areas of the network; the RAN, the core, the transmission, broadband, TV, our handsets, our probe data. And it’s how we are able to manage all that information to be able to come up with a coherent view, to help us both from an engineering viewpoint and an optimisation front, and also to help us from a customer perspective. How do we enable our care teams to fix issues quicker? identify issues quicker?
One of the key areas for me was that we needed to look at both our RAN and our Core. There have been numerous occasions where we have had a change in our Core and it may have an impact on our RAN network, but we weren’t able to identify it quick enough. Having all the data in one system, and having the reports set up in one system allowed us to capture those quicker, identify the issues quicker and fix them quicker, which is obviously a benefit to everybody.
So our journey. So initially we had our RAN data sources coming into the Zen system. We enriched that with Core data, our IMS core, our handset information, call trace, and signalling flows, and also brought in our probing. So by enriching our tools we were able to both fulfil the necessary requirement, which was ‘give us a better view of how our network is performing, allow us to see when we do optimisation, what actual impact is that having, is it actually working?’ I’m going to touch on one of the use cases shortly, where I’m going to show that historically, as a RAN engineer, if you up-tilted or you down-tilted the coverage, you assume that you improved it simply because your statistics on the RAN side told you so. That’s not necessarily the case, because there are obviously customers out there that may or may not loose coverage as a result.
So what have we achieved over the last year? One of the major areas for me was, right, ‘how do we measure the call quality and the signal strength when we make changes?’ In this graph, what we are showing here is a small little part of Kerry; Kerry is where Star Wars was filmed if anybody wants to know. What we have here is, we have a view of our EcNo and our RSCP. In the top right hand corner, here is what we would classify as poor service. So for us what we want to see when we make our changes, is we want to see everything flowing down in here (bottom left corner), so we can increase what we classify as good service, good radio conditions, and reduce the other areas. So we were also able to see that on a customer perspective, we can see that an individual customer could be serving on a cell up to 30 kilometres away. That might not necessarily be the most important cell for that customer to be on. So that sort of visibility allowed us to optimise the network, to improve the network, both from a statistical viewpoint, but also from a customer view point – give them a better service.
This is a very interesting use case that we had where we weren’t able to see when a customer did a circuit switched fallback from the 4G network - we weren’t able to see exactly where did they land? Where did they go on our 2G or our 3G network? By incorporating two different data sources, one a probe data and the other a GPEH, we were able to identify very quickly through this use case where customers were doing a circuit switched fallback from our 4G network to our 3G network, and the wrong cell, the right cell, the wrong area? And also exactly what was the success rate for circuit switched fallback. This is really powerful because it allowed us to optimise our 4G network and optimise our circuit switched fallback parameters to ensure that again, the customer has a great experience.
I touched on this earlier on, the important thing for us is to move our customers from the top right hand corner down to the bottom left hand corner. What this does, is that it gives us an idea that the network RAN performance is meeting our customer needs, that actually when we are doing our optimisation programmes, that we are actually improving the network, we are not just improving the traditional KPIs – this is actually an improvement of the network performance, and network experience.
One other area that we now use Zen to bring in is, how is our fixed network performing? So what we have done, we have brought in different data sources from our fixed areas, from our wholesale providers, from our own product, and have built a report that looks at the network performance on the fixed side versus our tickets. So from a customer care perspective we can see, if we have an increase in tickets, what is the impact on the customer? Do we have an increase in the number of disconnects, do we have an increase in broadband speed issues?
This is another use case that we wanted to try and develop, and we have done successfully, where we actually look at ‘how is our IP voice network performing?’ Again on our fixed network and our fixed services. So we are able to come up with a view of a good measured ratio versus a critical measured ratio, in terms of, I’s our IP network performing as it should be?’
The next steps for us is to come up with a way of correlating all of this data, automating all of this data, and making it easy for all of the different users of the data to do their jobs efficiently and effectively – that’s the end game for us.
So as you can see, we can close the loop, we identify any future problems and any customers that may have issues going forward. We get an understanding from a Netperform smartphone application, by bringing that data source in we get an understanding of how is the actual handset is performing, and how is the network performing for those customers. We also include and develop a customer care portal. We have a customer care portal at the moment for our fixed side, and we want to develop a mobile side as well.
That essentially is the journey for Vodafone Ireland from an optimisation viewpoint. We have come a long way; we have a long way to go.